交互式多模型算法过渡过程的仿真分析

Interacting Multiple Model Algorithm in Transient Process

交互式多模型(IMM)是混合估计理论中比较优越的算法。混合系统假设模式的切换能够瞬时完成。然而由于量测噪声和滤波系统惯性的作用,混合系统的模式切换具有过渡过程,对于实际运动模式切换存在漏检与误报。本文利用Monte Carlo仿真分析了两模型IMM算法过渡过程的滤波残差的统计特性,研究结果对于分析和改进算法具有重要的意义。

Interacting Multiple Model Algorithm (IMM) has been demonstrated to be the most cost-effective algorithm in hybrid estimation. But its performance degrades when the target real model has changed. The transient process of IMM is proposed to account for the performance degradation. Simulation results of the transient process are analyzed in this paper.